haidrrrry

haidrrrry / compose-kotlin-agent-skills

Public

Jetpack Compose & Kotlin AI agent skills for Cursor, Claude Code, Codex, Gemini & 27+ agents. Strict MVI · Kotlin 2.x K2 · Compose 2026 · CI-validated.

github.comhaidrrrrycompose-kotlin-agent-skills#readme agents-md ai ai-agents android android-development
15
0
89% credibility
Found May 21, 2026 at 15 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Markdown
AI Summary

This project is a knowledge library that teaches AI coding assistants how to write better Kotlin and Jetpack Compose code for Android apps. Instead of being a library or tool you install, it's a collection of markdown files that you give to your AI assistant. When your AI reads these files before writing code, it learns the right way to handle things like state management, animations, database access, and navigation — and learns what mistakes to avoid. The result is that your AI generates code that not only compiles but follows professional Android development practices from the start.

How It Works

1
🤔 You notice your AI writes buggy code

Your AI assistant generates Android code that compiles but has hidden problems like memory leaks and race conditions.

2
💡 You discover this skill kit

Someone tells you about a collection of instructions that teaches AI assistants how to write proper Kotlin and Jetpack Compose code.

3
📦 You download the knowledge pack

You copy the files into a special folder where your AI assistant can read them before writing code.

4
🔗 You connect it to your AI assistant

Following a simple guide, you point your AI to the new instructions — takes just a few minutes.

5
You start building with your AI
✅
Your AI uses proper patterns

It follows best practices like MVI architecture, atomic state updates, and lifecycle-aware flows.

🚫
Your AI avoids banned mistakes

It steers clear of antipatterns like GlobalScope, mutableStateListOf in ViewModels, and hardcoded strings.

🎉 Your app just works better

No memory leaks, no race conditions, no broken scrolling — your AI now writes production-ready Android code from the start.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 15 to 15 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is compose-kotlin-agent-skills?

This is a markdown-based knowledge kit that teaches AI coding agents how to write proper Jetpack Compose and Kotlin code. Instead of manually reviewing every suggestion your AI assistant makes, you give it this skill set and it learns the right patterns upfront. The kit covers strict MVI architecture, Kotlin 2.x compiler requirements, Compose 2026 APIs, and includes a banned-antipatterns table that explicitly prevents common mistakes like using `_state.value` instead of atomic updates, `collectAsState()` instead of lifecycle-aware collection, or `GlobalScope.launch` inside ViewModels. It works across 27 different AI agents including Cursor, Claude Code, Codex, Gemini, and GitHub Copilot.

Why is it gaining traction?

AI coding agents are now common in developer workflows, but they consistently generate Compose code that compiles yet violates Android best practices. This tool directly addresses that pain point by embedding guardrails into the agent itself. The hook is the banned-antipatterns table - developers immediately see value when their AI stops suggesting `mutableStateListOf` in ViewModels or hardcoded UI strings. The setup claims to take two minutes, and the modular structure lets agents load only relevant knowledge on demand.

Who should use this?

Android teams using AI coding assistants in production Kotlin/Compose codebases. Mobile developers tired of rejecting AI-generated code that breaks scroll performance or leaks memory. Teams maintaining strict architecture standards who want their AI assistant to enforce those standards automatically rather than through manual review cycles. Not useful if you don't use AI coding agents or work outside the Jetpack Compose ecosystem.

Verdict

The concept is solid and the execution - comprehensive coverage of 13 topics with CI validation - shows thoughtful design. However, the credibility score of 0.9% and only 15 stars indicate this is early-stage with minimal community validation. Try it if you're already deep in Compose development with AI agents, but treat it as an experimental addition to your workflow rather than a mission-critical tool until it accumulates more real-world usage and community feedback.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.